Graph rigidity, cyclic belief propagation, and point pattern matching

A recent paper \cite{CaeCaeSchBar06} proposed a provably optimal, polynomial time method for performing near-isometric point pattern matching by means of exact probabilistic inference in a chordal graphical model. Its fundamental result is that the chordal graph in question is shown to be \emph{globally rigid}, implying that exact inference provides the \emph{same} matching solution as exact inference in a complete graphical model. This implies that the algorithm is optimal when there is no noise in the point patterns. In this paper, we present a new graph which is also globally rigid but has an advantage over the graph proposed in \cite{CaeCaeSchBar06}: its maximal clique size is smaller, rendering inference significantly more efficient. However, this graph is not chordal and thus standard Junction Tree algorithms cannot be directly applied. Nevertheless, we show that loopy belief propagation in such a graph converges to the optimal solution. This allows us to retain the optimality guarantee in the noiseless case, while substantially reducing both memory requirements and processing time. Our experimental results show that the accuracy of the proposed solution is indistinguishable from that of \cite{CaeCaeSchBar06} when there is noise in the point patterns.     

Processing of tomato: impact on in vitro bioaccessibility of lycopene and textural properties

BACKGROUND: Human studies have demonstrated that processing of tomato can greatly increase lycopene bioavailability. However, the difference between processing methods is not widely investigated. In the current study different thermal treatments of tomato were evaluated with regard to their impact on in vitro bioaccessibility and retention of lycopene and β‐carotene as well as textural properties. Thermal treatments used were low (60 °C) and high (90 °C) temperature blanching followed by boiling. RESULTS: Lycopene was relatively stable during thermal treatment, whereas β‐carotene was significantly (P < 0.05) reduced by all heat treatments except for low temperature blanching. In vitro bioaccessibility of lycopene was significantly increased from 5.1 ± 0.2 to 9.2 ± 1.8 and 9.7 ± 0.6 mg kg^?1 for low and high temperature blanching, respectively. An additional boiling step after blanching did not further improve lycopene bioaccessibility for any treatment, but significantly reduced the consistency of low temperature treated samples. CONCLUSION: Choice and order of processing treatments can have a large impact on both lycopene bioavailability and texture of tomato products. Further investigations are needed, but this study provides one of the first steps towards tomato products tailored to optimise nutritional benefits. Copyright © 2010 Society of Chemical Industry     

Optimization of Robust Loss Functions for Weakly-Labeled Image Taxonomies

The recently proposed ImageNet dataset consists of several million images, each annotated with a single object category. These annotations may be imperfect, in the sense that many images contain multiple objects belonging to the label vocabulary. In other words, we have a multi-label problem but the annotations include only a single label (which is not necessarily the most prominent). Such a setting motivates the use of a robust evaluation measure, which allows for a limited number of labels to be predicted and, so long as one of the predicted labels is correct, the overall prediction should be considered correct. This is indeed the type of evaluation measure used to assess algorithm performance in a recent competition on ImageNet data. Optimizing such types of performance measures presents several hurdles even with existing structured output learning methods. Indeed, many of the current state-of-the-art methods optimize the prediction of only a single output label, ignoring this ‘structure’ altogether. In this paper, we show how to directly optimize continuous surrogates of such performance measures using structured output learning techniques with latent variables. We use the output of existing binary classifiers as input features in a new learning stage which optimizes the structured loss corresponding to the robust performance measure. We present empirical evidence that this allows us to ‘boost’ the performance of binary classification on a variety of weakly-supervised labeling problems defined on image taxonomies.     

Interaction of erythrocytes with magnetic nanoparticles

Internalization of biocompatible magnetic nanoparticles by red blood cells (RBCs) is a key issue for opportunities of new applications in the biomedical field. In this study, we used in vitro tests to provide evidences of magnetic nanoparticle internalization by mice red blood cells. The internalization process depends upon the nanoparticle concentration and the nanoparticle hydrodynamic radii. The cell internalization of surface-coated maghemite nanoparticles was indirectly tracked by Raman spectroscopy and directly observed using transmission electron microscopy. The observation of nanoparticle cell uptaking using in vitro experiments represents an important breakthrough for the application of nanomagnetism in diagnosis and therapy of RBC-related diseases.     

Stability and Quality Parameters of Probiotic Cantaloupe Melon Juice Produced with Sonicated Juice

The use of sonicated melon juice as substrate for Lactobacillus casei growth in cantaloupe melon juice and its effect on product quality were investigated. The survival of L. casei in the fermented juice, the product color, the post-acidification and other quality parameters were evaluated in the refrigerated product (4 °C) for a period of 42 days. Storage pH, color, sugar content, lactic acid concentration, biomass and viable cells count were determined. The caloric value of the product was also evaluated. During the refrigerated storage period, acidification and a slight reduction in the number of viable cells were observed. However, a sufficient number of viable cells were observed to guarantee the beneficial properties from probiotic consumption (>8 log CFU?ml^?1) throughout the storage period, for both stocked samples (with and without sugar addition). The product color was also preserved during storage. The developed product resulted in a low caloric product. The caloric value was reduced during the storage period due to sugar consumption by the remaining microorganisms.     

Storage Stability and Acceptance of Probiotic Beverage from Cashew Apple Juice

The stability of nonfermented cashew apple juice (control) and the fermented juice with Lactobacillus casei NRRL B-442 (sweetened and nonsweetened samples) was investigated along the storage at 4 °C for 42 days. The viability of the probiotic bacteria, sugars, lactic acid, and vitamin C content besides color, antioxidant activity, enzyme activity, and sensory characteristics of probiotic cashew apple juice was evaluated. It was observed that viable cell counts increased in the probiotic cashew apple containing sucrose (8 % w/v) along the storage period. Viable cell counts were always higher than 8.00 Log CFU/ml throughout the storage period. Ascorbic acid loss was higher in the nonfermented juice (40 %) compared to the fermented juice (23 %). The same behavior was observed for antioxidant activity and total polyphenolic compounds content, conferring nutritional benefits to this functional food. Browning reactions and nutritional breakdown caused by enzymes were minimized in the fermented samples during storage. In the fermented sample, higher reduction (>70 %) in the polyphenol oxidase enzyme activity was observed. During storage, the increase in the chroma values from 3.2 to 5.0 indicated that the yellow color became more vivid. Sensory analysis of the fermented cashew apple juice revealed that the product was well-accepted, with acceptance percentages above 80 % for the sweetened juice at the end of the storage period.     

Fast matching of large point sets under occlusions

The problem of isometric point-pattern matching can be modeled as inference in small tree-width graphical models whose embeddings in the plane are said to be ‘globally rigid’. Although such graphical models lead to efficient and exact solutions, they cannot generally handle occlusions, as even a single missing point may ‘break’ the rigidity of the graph in question. In addition, such models can efficiently handle point sets of only moderate size. In this paper, we propose a new graphical model that is not only adapted to handle occlusions but is much faster than previous approaches for solving the isometric point-pattern matching problem. We can match point-patterns with thousands of points in a few seconds.     

Graphical models and point pattern matching

This paper describes a novel solution to the rigid point pattern matching problem in Euclidean spaces of any dimension. Although we assume rigid motion, jitter is allowed. We present a noniterative, polynomial time algorithm that is guaranteed to find an optimal solution for the noiseless case. First, we model point pattern matching as a weighted graph matching problem, where weights correspond to Euclidean distances between nodes. We then formulate graph matching as a problem of finding a maximum probability configuration in a graphical model. By using graph rigidity arguments, we prove that a sparse graphical model yields equivalent results to the fully connected model in the noiseless case. This allows us to obtain an algorithm that runs in polynomial time and is provably optimal for exact matching between noiseless point sets. For inexact matching, we can still apply the same algorithm to find approximately optimal solutions. Experimental results obtained by our approach show improvements in accuracy over current methods, particularly when matching patterns of different sizes.